6v6 octet

[Andy Le Blanc]

Working on half a notion, looking at the bone pile in the shop.

Can you run an octet of 6v6 on a 2k twin OPT?

The half notion gets worse, but this is where it could be interesting.

I noticed that old wards/airline sometimes ran the same OPT for either a pair 6l6 or quad 6v6... they specified the same primary resistance.

[TUBEDUDE]

Much love for V4 amps. A V8, please do! Sound clips please.

[Firestorm]

I think that would sound superb. Depending on voltage of course.

[Andy Le Blanc]

worth a try eh...

[Andy Le Blanc]

generic twin iron gives you 450ma to play with,.. a deluxe is 120ish ma. seems close.

The rest of the half notion was to self bias, on series heaters of the pre amp tubes.
Again following wards/airline, they did this on pair 6l6 and quad 6v6.
I've done this and it works, the only drawback is the wait for the controlled heater warm up period.

I checked the math around bias for a pair 6l6 and that checks two.

Extrapolating up to 8 6v6, and a series string under each leg of the push pull circuit..

Its really being a cheap skate, dressing an amp with as little as possible, using up the bone pile in a interesting way.

[Stevem]

I don't about that for gig useage as the more power valves you have the more chances you have for a failure that will shut you down, a lot of wasted idle heat too, just in regards to all the heaters!
And on the other side of things if you run all those tubes light enough you will not be grinding them to get the over drive nature they lend to the mix.

[martin manning]

I don't about that for gig useage as the more power valves you have the more chances you have for a failure that will shut you down, a lot of wasted idle heat too, just in regards to all the heaters!
And on the other side of things if you run all those tubes light enough you will not be grinding them to get the over drive nature they lend to the mix.

Four 6L6's need 3.6A for the filaments and eight 6V6's need the same 3.6A! Pa max is 120W for four 6L6's vs. 96W for eight 6V6's! Total heat dissipation at idle is actually about 15% lower if they are run at the same percentage of Pa max!

[gui_tarzan]

I don't think it would be feasible for gigging but imagine that sound in a studio?

[Stevem]

You could do it for gigging, I would want a blow resistor in the plate circuit like the SVT and some other amps make use of!

[Firestorm]

I imagine you could do something "improper" like Marshall did with the 100 watts. If you double or quadruple the tubes and and halve or double everything as you're supposed to, it should sound the same, just more power. But if you don't adjust the time constant and the filtering, etc., you get something else.

[Stevem]

?

[Firestorm]

?

If you compare 50W Marshalls with 100W versions, there should have been specific changes to the supply filtering, the feedback loop, and the PI to compensate for a doubling of transconductance, plate current and input capacitance. The changes that were made over time seem almost random, so you get something other than a straight doubling of a JMP50. A quartet of 6V6s might not load down the PI too badly, but and octet would, so one might make changes there as well. Or not. I'm just saying that doubling or quadrupling output power isn't quite as simple as it first appears.

[Phil_S]

I am swimming in 6AQ5's, the little brother of the 6V6. I have often thought about building an 8x 6AQ5 amp just to see what it will do. I think these would need Va <= ~310VDC and likely a bit less. So, that wouldn't fill the bill for what's being discussed here. Nevertheless, I'm really very happy to see this thread, Andy, and am looking forward to learning what you find out! I think you've got the makings of a terrific cream machine!

[Andy Le Blanc]

the notion was to be as primitive as possible, if the loading works...

I've been inside enough rigs and taken the time to check out designs in a practical way as well.

SO... very simply, if the loading works, I 've got other design notions to explore as well.

I know that biasing on series heaters work, as I said the only real drawback is the controlled heater warm up time, its like waiting for the old TV to warm up, If anybody is old enough to remember that you'll know.
Each side of the P-P circuit will biased on the heaters of two series dressed 12a-7 type tube, I know it works, but I haven't personally tried a P-P with each leg separately bias, Matchless and a few others tried that, sounds fun to me.

Another thing I want to try is the very old fashioned approach of tying all the screen grids together as one.

I know this flies in the face of design considerations for an octet of tubes.

I figured to try, even though all the screens are tied to a common source, each can be dressed with a bleeder to ground to obtain enough current flow and voltage drop to regulate the common screen voltage.

I want to explore this, I had a clients amp that was old enough to have this design approach, and even though it was a small amp it had a voice and volume that was unreal.

Its going to have a primitive paraphrase invertor, it sounds like it might not have enough drive for the power side, but I think my silly simple design might already have a solution.

Biasing with the tube heaters gives me 4 pre-amp tubes to play with.
So... I was thinking of doubling the valves of the inverter, each side of the invertor will be a dual triode, geek fun why not.

And the pre stage I wanted to try paralleling two tubes, or 4 triode sections.
I saw this once in an old JAN radar manual.

Now I know its silly, but I've got enough crap around the shop to try it.
And if all else fails theres enough resources in the build to modify the design if I'm completely full of crap.

Got to have some fun, and theres always more to explore.

so theres several points.

loading, biasing, screens

and the effects of creatively paralleling tubes in the pre amp.

all to end up with a very simple rig with an octet of 6v6.

half a notion...